Installing Slackware Linux Part 2 - setup
Now
that we have our Linux partitions created, at the root prompt we can
type setup
This
is the main setup menu. You can read the help if you like, but you
can just skip down to ADDSWAP unless you need to remap your
keyboard for some reason. Use the arrow keys to navigate, and enter
to select.
It
will detect your swap partition for you, format it (mkswap) and
activate it (swapon)
Note: The hard disk devices in these
screenshots are /dev/sda. Don't pay any attention to that, it's just
because I took these screenshots from within a virtual machine. It
emulates disks as scsi devices. Just know that's not the disk we
partitioned in the examples above, so there's no confusion.
After
completing a step, setup automatically takes you to the next step in
sequence. Next, is to select the target partitions. Here is where we
choose our root partition, and then choose mount points for the other
partitions.
This
is where we select our root partition (/). Following our partitioning
example, that would be /dev/hda2.
Now
it will prompt you to format the partition. I would choose to check
for bad blocks while it's formatting.
Choose
your desired filesystem. I like to use ext2 because it's a simple
filesystem that's well matured, but you may want to choose ext3 or
reiserfs to have a journaling filesystem.
Now
it prompts to choose the inode density for the filesystem. Just hit
enter to go with the default of 4096 unless you know what you are
doing, and specifically why you want to do it.
If you just
created a root partition and swap, you are done formatting now. If
you created other partitions, they must now be selected, formatted
and assigned mount points.
Swap
doesn't show up in this list.
We
are mounting this partition as /usr.
Continuing on, we are
prompted to select, assign mount points and format the rest of our
partitions in the same manner.
When
finished, a summary is displayed
In
the next step, you will be prompted to select the source
media.
Hit
enter to choose a Slackware CDROM, and it should detect it
automatically.
In the next step, we are prompted to select
package categories.
These
govern which series of packages will be installed on the system. By
default, all categories are selected except KDEI (KDE i18N
internationalization). If you're just going to be using
English/Western charsets you don't need to install KDEI.
For
your first time installing Slackware, I recommend leaving all package
categories enabled. You can just choose OK here.
Next, we are
prompted to choose the "prompt mode", that is, the degree
of interaction for installing packages.
Full,
installs all packages in the categories you've selected, without
prompting. This is what I recommend for your first Slackware install.
Install everything, and you can easily remove packages you don't want
later after you get a feel for things. I do know what I'm doing, but
this is the option I normally use. It's just easier.
Newbie
prompts for each package as they are being installed. I do not
recommend this, as it is quite tedious. Also, you may not know what
you want/need yet.
Menu is a bit better, as it lets you
choose groups of related things.
Expert. If you know
what you are doing, the expert prompt mode is an excellent way to
choose exactly which packages you want installed on the system, prior
to package installation. This really is good, it's not terribly
confusing like similar package installation modes in other
distributions.
The custom/tagfile options use tagfiles
to automate a custom package selection. I've never used them. This
would be handy if you were wanting to roll out the same installation
on several machines though.
Choose full and watch the
packages install non-interactively. It won't take very long, even on
a relatively slow machine.
At
some point during the package installation, you will be prompted to
insert the second CD.
When the package installation stage
completes, you are prompted to choose a kernel.
I
recommend the CDROM option, and choosing the same kernel that you
chose at the initial boot prompt when you booted with the Slackware
CD. It got you this far. (If you just hit enter to accept the
default, it's called bare.i)
Because
I did these screenshots in a virtual machine that uses scsi emulation
for the virtual disks, I had to choose scsi.s. On an IDE
system, you probably either want bare.i or bareacpi.i (warning: acpi
can cause boot problems if your BIOS implementation of ACPI doesn't
jibe... this is why I recommend using the same kernel you chose at
the initial boot prompt)
Next, you are prompted to create an
emergency boot disk.
I
highly recommend taking the time to create this disk, for it can be
used to start the distribution if anything ever happens to your boot
loader. You will be able to easily fix it, if you can start the
system using this boot floppy.
You will now be prompted to
create a symbolic link for your modem device.
If
you have a modem, you can do that here. Saying "no modem"
doesn't mean you can't use a modem, you can create the /dev/modem
symbolic link later, or just use the appropriate device (e.g.
/dev/ttyS1 for COM2)
Next, you will be prompted to enable the
hotplug system. If you have such devices, say Yes, otherwise No is a
good idea.
As
you can see, it's possible for it to cause problems on some systems.
Note the information on how to get out of the trap if it happens to
you.
Now we are prompted to install the LILO
bootloader.
You
will most likely want to choose simple here. Choosing expert,
will result in lilo not behaving as you expect and you'll have to
manually edit the lilo.conf file (or run liloconfig from within the
OS) to get the desired functionality back (e.g. it won't even prompt
you to select an operating system). If you choose to skip the
installation of lilo altogether, then you will only be able to boot
into your Slackware system using the boot floppy that you created in
the previous steps.
Next, you are prompted to choose the VGA
(display) mode of your console, either standard VGA, or one of the
VESA framebuffer display modes. The reason this is in the lilo
configuration, is because the boot loader passes these parameters to
the kernel on boot.
It
is nice to have a framebuffer console for when you're not running
XFree86, but if the framebuffer mode you've chosen doesn't work well
with your display hardware, you could end up with an unusable display
(until you fix it of course... you could boot with your boot
floppy).
Consider choosing standard for now, to use
standard VGA. You can change this parameter in your /etc/lilo.conf
file later. If you're always going to be using the XFree86 GUI
environment, it's not going to matter much anyways.
You are
now prompted to enter any extra boot parameters, that lilo is to pass
to the kernel.
He
gives one very common example of why you might need to do this: If
you have an IDE CD Writer. In the 2.4 kernel series, CD writing uses
SCSI emulation and the kernel must know which drive is to use that
mode, if both IDE-CD Support and IDE-SCSI support are to be loaded in
the running kernel. The example of hdc, is for a secondary master.
Use hdd if your writer is secondary slave.
Next, you are
prompted to choose the destination for installing LILO. You will
almost certainly want to choose MBR (unless you know what you
are doing)
He
says "possibly unsafe" because there are a few situations
where writing to the master boot record is indeed unsafe. For
example, if your bios doesn't support the capacity of the drive, and
you have translation software installed (e.g. "MaxBlast" or
"EZBios"). Another reason it could be unsafe is, if you are
using another boot loader (e.g. System Commander, or Boot Magic).
Also, before you ever write to the MBR (installing pretty much any
OS), you must ensure that bios level MBR protection is disabled.
(a.k.a boot virus protection, or "Trend ChipAway").
Installing LILO to the MBR, is the most common way that it is used
and it is normally quite safe and can be used to start your Windows
operating systems as well.
The "Root" option, to
install LILO to the superblock of your root partition, is mainly
useful if you intend to use another boot manager to invoke
LILO.
Next, you are prompted to create a symbolic link for
your mouse.
Even
if you don't intend to use gpm, it's still useful to have a correct
/dev/mouse symbolic link. This way you can just specify that device
when you configure XFree86 after the OS is installed. I choose imps2
for my Logitech ps/2 wheel mouse.
I
don't have much use for this (it's got nothing to do with using a
mouse in the GUI), but if you wish to have mouse support at the
console, you can load gpm at boot time.
At this point, you
will be asked if you want to Configure your network. If you
only have dial up networking, and don't even have a NIC, you can say
No to that question for now, and you'll be prompted to configure your
clock, timezone and set a root password. Alternatively you can
proceed, and choose loopback. That is really what you should do, as
then at least you'll set a hostname for the machine.